Code optimization
Description:
Minor tweaks throughout the optimized regular hyperslab code to increase
speed. This set of improvements increase the benchmark time from taking
~5.46 seconds to ~4.50 seconds, or around a 20% further speedup.
Platforms tested:
Solaris 2.6 (baldric)
Feature symmetry
Description:
A while ago I needed to get the 'type' of data being accessed during writes
to the VFL driver, so I put in code to get the information down there.
Albert asked for the same information during reads, so I've added that in.
Tested:
FreeBSD 4.1.1 (hawkwind)
Bug Fix
Description:
When parallel I/O is turned on, there were some macros used in the H5D
routines which poked around in the H5F_t structure. This breaks the
privacy of that structure and ties the H5D code too tightly to the H5F_t
struct.
Solution:
Added a small function to retrieve the the value (driver_id) needed from
the H5F_t function.
Platforms tested:
Eyeballed only, Albert needs this right away...
Maintainance & performance enhancements
Description:
Re-arranged header files to protect private symbols better.
Changed optimized regular hyperslab I/O to compute the offsets more
efficiently from previous method of using matrix operations.
Added sequential I/O operations at a more abstract level (at the same level
as H5F_arr_read/write), to support the optimized hyperslab I/O.
Platforms tested:
Solaris 2.6 (baldric) & FreeBSD 4.1.1 (hawkwind)
Rearrange code
Description:
The data sieve buffering code for contiguously stored datasets was
wedged in the H5F_arr_read/H5F_arr_write routines.
Solution:
Created a new H5Fcontig.c to hold I/O routines for contiguously stored
datasets (like H5Fistore.c for chunked dataset I/O routines) and moved
data sieving code into those routines.
Platforms tested:
Solaris 2.6 (i.e. baldric)
Implemented new feature
Description:
Added data sieve buffering code to raw I/O data path. This is enabled for
all the VFL drivers except the mpio & core drivers. Also added two new
API functions to control the sieve buffer size: H5Pset_sieve_buf_size() and
H5Pget_sieve_buf_size().
Platforms tested:
Solaris 2.6 (i.e. baldric)
has turned off error reporting in the library.
Also, changed the way property lists are managed and closed so that they
can be automatically closed when the library terminates.
to double-check things. I've turned them back on again now. I also changed
the internal representation of a few struct fields to be float instead of
double, since the HP/UX 10.20 compiler was having problems with the alignment
of the doubles.
but I fixed lots of misc. compiler warnings in other code and also tracked down
the memory overwrite bug that was causing the development branch to core dump
on most machines.
----------------------
./src/H5D.c
./src/H5F.c
./src/H5FDmpio.c
./src/H5Fistore.c
./src/H5Fprivate.h
./src/H5Smpio.c
The `driver_id' for a file was accidently put in two
structs. I removed it from the H5F_file_t struct since it's
really an attribute of the VFL stuff.
More careful incrementing/decrementing the driver ID to fix a
memory leak.
./src/H5P.c
Rewrote H5Pcreate() in terms of H5P_copy() of a default
property list. This fixes some referencing counting bugs.
----------------------
./MANIFEST
./src/H5FDmulti.c [NEW]
./src/H5FDmulti.h [NEW]
./src/Makefile.in
./src/hdf5.h
The split driver was reimplemented as a more general "multi"
driver which is capable of splitting data into multiple files
like the family driver except the partioning is done by memory
usage type instead of address. The H5Pset_fapl_split()
function just calls H5Pset_fapl_multi() with arguments which
prepare to split the address space into two files: meta and
raw data.
This is the first version. I plan to allow the open() call to
relax a bit which would allow one to open an hdf5 file when
only the meta-data file is present. This would allow a very
large file to be split and stored on tape and the relatively
small meta file to be mirrored on disk to allow limited
browsing of the file (any request for raw data would fail).
./src/H5private.h
./src/H5F.c
./src/H5FD.c
./src/H5FDprivate.h
./src/H5FDpublic.h
./src/H5FDcore.c
./src/H5FDfamily.c
./src/H5FDmpio.c
./src/H5FDsec2.c
Added the ability for a file driver to store information in
the superblock which would be needed if the file were opened
again later for reading. The format is driver-defined which
allows users to extend it however they like.
./doc/html/H5.format.html
Added information about the new driver information block of
the superblock. This is where file drivers store information
they need in order to reopen the file later.
./src/H5F.c
./src/H5Fprivate.h
./src/H5FD.c
./src/H5FDprivate.h
./src/H5FDpublic.h
./src/H5FDcore.c
./src/H5FDfamily.c
./src/H5FDmpio.c
./src/H5FDsec2.c
./src/H5Fistore.c
./src/H5R.c
The file access properties and the file access property list
were decoupled, which allows the property list to more cleanly
contain properties for various levels of the file and which
allows the property list to be modified more cleanly when
opening files.
./src/H5.c
./src/H5FDpublic.h
Removed H5FD_MEM_META and H5FD_MEM_GROUP since they're never
used.
./src/H5D.c
Changed the way we detect the MPIO driver in all these special
cases.
./src/H5F.c
./src/H5Rpublic.h
./test/tfile.c
The default file sizeof(offset) was changed to be a function
of haddr_t instead of hsize_t.
THE H5RPUBLIC.H DEFINITIONS WILL HAVE PROBLEMS IF THE USER
CREATES A FILE WITH NON-DEFAULT OFFSET AND SIZE SIZES!
./src/H5F.c
Fixed an uninitialized memory access bug in file closing
related to the VFL.
./src/H5T.c
./src/H5Tpublic.h
Added an H5T_NATIVE_HADDR predefined datatype which
corresponds to the `haddr_t' type.
./test/Makefile.in
Reformatted long lines.
./test/big.c
./test/cmpd_dset.c
./test/dsets.c
./test/dtypes.c
./test/extend.c
./test/external.c
Removed the H5F_ACC_DEBUG flag from file creation/open calls.
./test/big.c
Plugged a memory leak.
./test/h5test.c
Added support for the `multi' driver.
Removed #warning about not having the stdio driver. Plans are
to not implement it since the sec2 driver serves the same
purpose and testing didn't show any difference in execution
times between the two.
----------------------
This extensive change is the virtual file layer implementation. I've
ported and tested the sec2, family, and core drivers and only ported
the mpio driver (Albert will test it). So if you need MPIO I would
recommend sticking with the previous version for a while.
You will get a few compile warnings about split and stdio drivers not
being implemented and possibly tracing information not inserted in
some of the drivers. You can safely ignore them but I plan to fix
them.
I'm still working on the split driver because I just realized that it
needs a part of the VFL that isn't written yet.
Documentation is being updated also because there were some minor
changes (mostly just name changes). It should be available on my web
site later this week.
./MANIFEST
./src/Makefile.in
./src/hdf5.h
./src/H5Flow.c [REMOVED]
./src/H5Fstdio.c [REMOVED]
./src/H5Fsec2.c [REMOVED]
./src/H5Fsplit.c [REMOVED]
./src/H5Fmpio.c [REMOVED]
./src/H5Ffamily.c [REMOVED]
./src/H5Fcore.c [REMOVED]
./src/H5MFpublic.h [REMOVED]
./src/H5FD.c [NEW]
./src/H5FDcore.c [NEW]
./src/H5FDcore.h [NEW]
./src/H5FDfamily.c [NEW]
./src/H5FDfamily.h [NEW]
./src/H5FDmpio.c [NEW]
./src/H5FDmpio.h [NEW]
./src/H5FDprivate.h [NEW]
./src/H5FDpublic.h [NEW]
./src/H5FDsec2.c [NEW]
./src/H5FDsec2.h [NEW]
Removed/added files for virtual file layer.
./bin/trace
./src/H5.c
Removed unused public datatypes and added new VFL public
datatypes.
Changed an error message.
./config/BlankForm
./config/dec-flags
./config/gnu-flags
./config/hpux10.20
./config/hpux9.03
./config/irix5.x
./config/irix6.x
./config/solaris2.x
./config/unicosmk
Removed the H5F_OPT_SEEK and H5F_LOW_DFLT constants from the
configuration since they're no longer applicable. The default
file driver is always the sec2 driver and it always optimizes
calls to lseek() or lseek64().
./config/depend.in
C preprocessor errors generated during automatic dependency
building are sent to /dev/null to prevent them from appearing
twice in the make output.
./src/H5AC.c
./src/H5B.c
./src/H5D.c
./src/H5F.c
./src/H5G.c
./src/H5Gent.c
./src/H5Gnode.c
./src/H5HG.c
./src/H5HL.c
./src/H5O.c
./src/H5Oattr.c
./src/H5Odtype.c
./src/H5Oefl.c
./src/H5Oshared.c
./src/H5T.c
./src/H5detect.c
./test/ohdr.c
Changed H5F_ADDR_UNDEF to HADDR_UNDEF to be more consistent
with the `haddr_t' datatype which is now a public type.
./src/H5D.c
./src/H5P.c
./src/H5Ppublic.h
./src/H5Tconv.c
./test/cmpd_dset.c
./test/dsets.c
./test/overhead.c
./test/tselect.c
./test/tvltypes.c
The H5P_DATASET_XFER constant was changed to H5P_DATA_XFER
because the properties apply to all types of I/O operations,
not just datasets.
./src/H5B.c
./src/H5Bprivate.h
./src/H5D.c
./src/H5Dpublic.h
./src/H5F.c
./src/H5Farray.c
./src/H5Fistore.c
./src/H5Fprivate.h
./src/H5Fpublic.h
./src/H5Gnode.c
./src/H5Gpkg.h
./src/H5HG.c
./src/H5HL.c
./src/H5O.c
./src/H5R.c
./src/H5Sall.c
./src/H5Shyper.c
./src/H5Smpio.c
./src/H5Spoint.c
./src/H5Sprivate.h
./test/big.c
./test/h5test.c
./test/istore.c
./testpar/t_dset.c
./testpar/t_file.c
./tools/h5debug.c
./tools/h5ls.c
Modified to work with the virtual file layer by calling H5FD_*
functions instead of H5F_low_* functions and by passing file
access and data transfer properties by object ID instead of
pointer.
Changed H5D_transfer_t to H5FD_mpio_xfer_t since the
COLLECTIVE vs. INDEPENDENT transfer mode is specific to the
MPIO file driver.
Moved MPIO-specific stuff into the MPIO driver.
./src/H5B.c
./src/H5D.c
./src/H5Fprivate.h
The H5F_mpio_* private functions were renamed and placed in
the H5FDmpio driver except those which appeared in H5Smpio.c.
./src/H5E.c
./src/H5Epublic.h
Added major error number H5E_VFL for virtual file layer
related errors.
./src/H5F.c
./src/H5Fprivate.h
Changed the logic that controls whether the boot block is
written. Instead of assuming that the first call to write the
boot block is only to allocate space, I've added a function
argument which makes this explicit.
Changed the way files are compared so that a driver-defined
comparison function can be called. Files which belong to
different drivers are always considered different.
Removed H5F_driver_t since file drivers are now identified by
object ID instead of a special non-user-extendible datatype.
Removed all the hard-coded low-level file properties which
have been replaced by the various file drivers.
./src/H5I.c
./src/H5Iprivate.h
Added the H5I_inc_ref() which was removed a few months ago
since we finally have a use for it.
./src/H5Ipublic.h
Added the H5I_VFL object ID type to identify file drivers in
the virtual file layer.
./src/H5MF.c
./src/H5MFprivate.h
Moved all the allocation/deallocation code into the virtual
file layer which allows file drivers to override much of it.
./src/H5P.c
./src/H5Ppublic.h
Moved file driver-specific code into the various file driver
files.
The H5Pcopy() and H5Pclose() functions make calls into the
virtual file driver to manage the memory for driver-specific
file access and data transfer properties.
./src/H5private.h
./src/H5public.h
The `haddr_t' type is now public.
./test/tfile.c
Added a few more comments.
final parameter (so that VL datatypes have a way to pass in the custom memory
allocation routines).
Fixed a conversion bug when VL fields were part of a compound datatype that
was causing the no-op conversion routine to be used instead of the vlen routine.
Added the H5Pset_vlen_mem_manager and H5Pget_vlen_mem_manager routines to
allow users to provide their own custom memory allocation routines for VL data
read in from the file and reclaimed with H5Dvlen_reclaim.
Finished coding on H5Dvlen_reclaim so it works now.
----------------------
Remove changes from CVS
./bin/release
Added a `--nocheck' switch which causes the script to not
check the contents of the MANIFEST file against CVS. This is
sometimes useful when you need to make a quick snapshot but
the MANIFEST file is not quite up to date.
./src/H5D.c
Removed warnings for unused variables
./src/H5Fprivate.h
Removed the WIN32 definition for `uint' and changed the data
type for `eof_written' from `uint' to `uintn'. Shouldn't this
really be `hbool_t'?
./src/H5Odtype.c
./src/H5T.c
./src/H5Tconv.c
./src/H5Tpkg.h
./src/H5Tpublic.h
./test/dtypes.c
./doc/html/H5.format.html
Added support for bitfields and opaque data types.
H5F.c:
H5Flow.c:
H5Fprivate.h:
Joined work by Quincey, Robb and me to eliminate the initial writes
of BOOTBLOCK when the file is first created. Also, eliminate the
repeatedly encoding of BOOTBLOCK in the flushing code. Introduced
the eof_written flag to indicate when it is not necessary to go
do a dumb write at eof.
(But something is not right if the userblock is not zero.)
H5Fmpio.c:
Added code to allow setting the MPIO debug output mask via
the environment variable H5F_mpio_Debug.
----------------------
./tools/h5ls.c
Added a `--address' (`-a') switch which causes h5ls to display
file addresses for raw data. For contiguous datasets it's just
a nice simple number, but for chunked datasets it's a list of
logical dataset coordinates, file addresses, filter masks, and
storage sizes.
Changed `--dump' switch to `--data'.
./src/H5D.c
./src/H5F.c
./src/H5Fprivate.h
Enhanced the indexed-storage B-tree iterator so it can dump
raw data addresses (and other info) to the standard error
stream.
Added H5Ddebug() so h5ls has a way to dump addresses for
datasets. I'm not sure what else this API function should do,
so I think we should discuss it before we document it. So far,
h5ls is the only thing that uses it, and we can easily change
that.
./src/H5Tconv.c
./test/dtypes.c
Finally had a chance to verify Paul's H5T_conv_s_s (general
string to string conversions) bug fixes and incorporate them
into H5T_conv_f_f (general floating-point to floating-point
conversions) and H5T_conv_i_i (general integer to integer
conversons). Thanks Paul.
./src/H5D.c
./src/H5S.c
./src/H5Sprivate.h
Added performance timers around data space read and write
callbacks. They were already there for the gather/scatter
callbacks.
The timings for read/write callbacks are displayed along with
gather/scatter when data space debugging is turned on.
./bin/iostats
Updated to print totals. Added a `--fast' option that doesn't
do any output except the totals and is much faster.
./bin/trace
Changed __unused__ to UNUSED to match source code.
./config/gnu-flags
Updated error message for pgcc. I've sent bug reports to the
pgcc people but the new version still has the same bug.
./configure.in
./config/conclude.in
./config/depend.in
Fixed dependencies for non-GNU makes when run in a directory
other than the hdf5 source tree.
Updated GNU `make dep' rules to copy the distributed
dependencies for non-GNU makes into the source tree when run
in some other directory.
----------------------
./src/H5B.c
./src/H5D.c
./src/H5F.c
./src/H5Farray.c
./src/H5Fcore.c
./src/H5Ffamily.c
./src/H5Fistore.c
./src/H5Flow.c
./src/H5Fmpio.c
./src/H5Fprivate.h
./src/H5Fsec2.c
./src/H5Fsplit.c
./src/H5Fstdio.c
./src/H5Gnode.c
./src/H5HG.c
./src/H5HL.c
./src/H5O.c
Changed H5F_block_read/write() to take a file transfer
property list instead of just a parallel I/O mode. This will
allow us to pass additional parameters to the I/O functions
without having so many arguments to worry about.
This fixes a bug I introduced last Friday for parallel HDF5
./src/H5B.c
Fixed decoding of B-tree keys when iterating through the
leaves of the B-tree. This fixes a bug for applications that are
adding new entries to groups and also listing the groups.
----------------------
./config/depend.in
Fixed automatic dependencies. We were storing dependencies for
*.o files instead of *.lo files after shared libraries were
added.
./config/gnu-flags
./config/linux-gnulibc1
Moved `-march=pentiumpro -mcpu=pentiumpro -malign-double' from
the linux file to this file and caused it to depend on the CPU
name. This fixes one of Elena's bugs.
./src/H5B.c
./src/H5Bprivate.h
./src/H5D.c
./src/H5Dprivate.h
./src/H5Dpublic.h
./src/H5F.c
./src/H5Farray.c
./src/H5Fistore.c
./src/H5Fmpio.c
./src/H5Fprivate.h
./src/H5Fpublic.h
./src/H5Gnode.c
./src/H5P.c
./src/H5RA.c
./src/H5Sall.c
./src/H5Shyper.c
./src/H5Smpio.c
./src/H5Spoint.c
./src/H5Sprivate.h
./src/H5Tpublic.h
./test/istore.c
Added an H5Dget_storage_size() function that reports the
amount of storage allocated for raw data in a dataset.
Changed H5D_xfer_* to H5F_xfer_* because these properties are
more general than datasets. This also allows some of the
lower-level I/O functions to get this information easier.
./src/H5S.c
./src/H5Sall.c
Added two new functions H5S_all_read() and H5S_all_write()
which are optimizations that copy data directly between file
and memory without having to go through the scatter gather
step. This knocks quite a bit of time off the I/O and
reading/writing entire datasets is a fairly common operation.
./tools/h5ls.c
Reports the logical size of data, the allocated size of data,
and the percent utilization.
./MANIFEST
Removed old pablo files, added new files. Snapshots should now
start to work again.
./src/H5D.c
./src/H5Fmpio.c
Removed two warnings signed vs. unsigned comparisons and check
for overflow.
./configure [REGENERATED]
./src/H5D.c
./src/H5O.c
Removed H5AC, H5B, and H5T from the default list of packages
to debug (because they're pretty expensive debugging), and
added H5O. Also fixed a bug for undefined variable in H5D when
H5S debugging is turned on but H5T debugging is turned off.
./config/conclude.in
Fixed installation of header files for building in a directory
other than the source directory. This fixes a bug where
H5config.h wasn't being installed.
./src/H5.c
./src/H5A.c
./src/H5D.c
./src/H5F.c
./src/H5G.c
./src/H5I.c
./src/H5Iprivate.h
./src/H5P.c
./src/H5R.c
./src/H5RA.c
./src/H5S.c
./src/H5T.c
./src/H5TB.c
./src/H5Tprivate.h
./src/H5Z.c
./src/H5detect.c
./src/H5private.h
Changed the way the library shuts down again. Now it handles
cycles between packages and isn't so sensitive to dependencies
between packages. A package might shut down only to be
restarted to process a request from some other package being
shut down. Loops are detected after 100 iteractions and the
shutdown is aborted with a message on standard error. This
also makes it a lot easier to debug.
./src/H5A.c
Fixed H5A_write() and H5A_read() so they pass a non-null
background buffer to the conversion functions. This is
necessary when an attribute has a compound data type.
./src/H5Flow.c
./src/H5Fprivate.h
./src/H5Fsec2.c
Reindented new Win32 stuff.
./src/H5Odtype.c
Fixed a bug when enumeration types are used in a compound data
type. The byte pointer wasn't incremented after the type
information was written.
./tools/h5ls.c
Compound data types display their total size because it's not
always obvious from looking at the members.
Scalar attributes show their space as `scalar' instead of
`{}'.
The index value is not printed for attributes that have only a
few values. Instead the word `Data:' is printed on the first
line of attribute data.
Named types display their data type only if verbose output was
requested.
added a new member to the H5F_search_t struct to store the path name
we needed this to be able to tell if a file is already open since NT does not
use the inode.
H5Flow.c:
Document previous changes (same changes committed by Robb first.)
H5Fprivate.h:
Change DECODE macros to do sign extension.
(Sign extension can be skipped if machines use the exact sizes.)
H5Sall.c:
Fixed a typo in return value.
H5private.h:
Fixed a typo and also changed the strategy of the int16_t and
uint16_t typedef. Now will use short as long as it is at least
2 bytes big.
----------------------
./src/H5F.c
./src/H5private.h
./src/H5Ipublic.h
./src/H5O.c
Fixed a rather nasty bug with file closing that caused the
file boot block to be updated incorrectly, effectively
truncating the file. The bug I fixed was triggered by:
1. Create a file, F
2. Open an object, X
3. Close file F
4. Reopen file F for read/write.
5. Create and close some objects
6. Close file F
7. Close library (exit).
Step 3 pended the close because object X is still open, but
the file ID was removed from the H5I_FILE ID group. Step 4
created a new file because it didn't see any matching file on
the H5I_FILE ID group. Step 5 extends the file. Step 6 writes
the new file boot block to disk. Step 7 closes object X and
completes the close from step 3, writing the old boot block
information to disk.
The new behavior is that step 3 moves the file from the
H5I_FILE group to the H5I_FILE_CLOSING group. Step 4 searches
both groups and finds the file. Step 5 extends the file using
the same H5F_file_t struct as step 3. Step 6 closes the H5F_t
struct opened in step 3 but not the H5F_file_t struct shared
by steps 1 and 3. Step 7 closes object X which closes the
H5F_file_t from step 1, flushing the boot block which was
shared by all steps.
./src/H5F.c
Added some bulletproofing to file reference counting and
removed comments which no longer apply. Added H5F_flush_all()
and H5F_close_all() which apply to all files.
./src/H5A.c
./src/H5D.c
./src/H5F.c
./src/H5G.c
./src/H5I.c
./src/H5Iprivate.h
./src/H5R.c
./src/H5RA.c
./src/H5S.c
./src/H5T.c
Added the new H5I_free_t data type to describe the function
type to be passed as the `free_func' argument to
H5I_init_group().
./src/H5I.c
Bulletproofed the object removal functions. Removed comments
which no longer apply. Changed global variable names so they
don't violate the naming scheme. Added H5I_debug() that prints
the contents of an ID group. Removed H5I_inc_ref() because it
isn't used. Reindented a couple of functions.
./src/H5.c
./src/H5G.c
./src/H5Ipublic.h
Changed H5I_MAXID to H5I_NGROUPS to better relect the fact
that it's the total number of valid ID groups.
./src/H5Shyper.c
Changed hyperslab offset arrays to signed quantities to get
rid of warnings on DEC cluster.
./src/H5Flow.c
./src/H5Fprivate.h
Changed the objno argument of H5F_addr_pack() to be unsigned
to get rid of warnings on DEC cluster.
----------------------
./src/H5Shyper.c
Changed hyperslab offset arrays to signed quantities to get
rid of warnings on DEC cluster.
./src/H5Flow.c
./src/H5Fprivate.h
Changed the objno argument of H5F_addr_pack() to be unsigned
to get rid of warnings on DEC cluster.
----------------------
./src/H5.c
Fixed more dependency problems in H5_term_library(). There was
a bug in the previous version that could cause the wrong EOF
marker to be written to the boot block under certain
circumstances. Hopefully this fixes it although I don't ready
access to a test case (Mark Miller will test it).
./src/H5F.c
./src/H5Fprivate.h
Added an H5F_close_all() that is similar to
H5F_term_interface() but which doesn't close the
interface. Files that don't have open object headers are
closed, others are delayed until all object headers close. All
files are flushed.
./src/H5ACprivate.h
./src/H5Bprivate.h
./src/H5Dprivate.h
./src/H5Eprivate.h
./src/H5Fprivate.h
./src/H5Gpkg.h
./src/H5Gprivate.h
./src/H5Gpublic.h
./src/H5HGprivate.h
./src/H5HLprivate.h
./src/H5Iprivate.h
./src/H5MFprivate.h
./src/H5MMprivate.h
./src/H5Oprivate.h
./src/H5Pprivate.h
./src/H5Ppublic.h
./src/H5RAprivate.h
./src/H5Sprivate.h
./src/H5Spublic.h
./src/H5Tpkg.h
./src/H5Tprivate.h
./src/H5Tpublic.h
./src/H5Vprivate.h
./src/H5Zprivate.h
./src/H5private.h
./src/H5public.h
Reindented after __DLL__ was added.
----------------------
./tools/h5tools.c
Strings are not converted to null-padding before being
printed; they are printed with whatever byte values appear in
the file.
./tools/h5ls.c
Now able to display attribute data type and data.
Added a `-f' or `--full' switch which causes the full name of
each object to be displayed instead of just the base name.
Added a `-r' or `--recursive' switch that recursively prints
the contents of groups, avoiding cycles.
More bulletproofing for non-printable characters in things
like object names, attribute names, and comment strings. We
don't want listing a file to send termal escape sequences
because it's sometimes possible to execute commands that way.
Since h5ls doesn't usually use quotes around object names we
must sometimes escape space characters.
External files are listed in a table to make the output less
confusing.
./tools/h5tools.c
./tools/h5tools.h
Changed h5dump() to h5dump_dset() and added h5dump_mem(). Also
make h5dump_fixtype() public.
./test/dtypes.c
Wrote some data to an attribute to test h5ls attribute
printing.
./src/H5ACprivate.h
./src/H5Apublic.h
./src/H5Dprivate.h
./src/H5Dpublic.h
./src/H5Epublic.h
./src/H5Fprivate.h
./src/H5Fpublic.h
./src/H5Gprivate.h
./src/H5HLprivate.h
./src/H5Oprivate.h
./src/H5Ppublic.h
./src/H5RApublic.h
./src/H5Rpublic.h
./src/H5Spublic.h
./src/H5Tpkg.h
./src/H5Tpublic.h
./src/H5Vprivate.h
./src/H5Zpublic.h
./src/H5private.h
./src/H5public.h
Reindented function prototypes after `HDF5API' was added. Also
rewrapped long lines.
./src/H5Flow.c
Added an `#ifdef WIN32' around an unused variable.
./src/H5api_adpt.h
Removed extra carriage returns inserted by "broken" operating
system.
./src/H5Dprivate.h
./src/H5Oprivate.h
./src/H5Vprivate.h
./src/H5private.h
Removed extraneous inclusion of H5api_adpt.h since it's
included in H5public.h which is included by everything.
./src/Makefile.in
Added H5api_adpt.h to the list of public header files to fix
broken `make install'.
----------------------
./INSTALL
Added warning that enabling debugging code can adversely
affect performance even when the debugging isn't turned on at
run-time. Performance testing shows that under certain
circumstances (like data type conversions of compound types)
the H5_timer functions, although only a few lines each,
contribute a fairly large percent to the execution time.
./src/H5T.c
./src/H5Tpkg.h
./src/H5Tpublic.h
Improved the H5Tunregister() function to make unregistering
more flexible. It takes the same arguments as H5Tregister()
but also accepts wild cards. All conversion functions that
match the H5Tunregister() search criteria are removed from the
global type conversion table.
The H5Tregister_hard() and H5Tregister_soft() were combined
into a single function called H5Tregister() which is the
counterpart to H5Tunregister(). A new `persistence' argument
was added to differentiate between the two types of conversion
functions.
The application is allowed to register a hard conversion
function for the no-op conversion path although the library
isn't obligated to call it (it usually does). This is mostly
for completeness, but the application might use it to help
determine if the raw data pipeline was able to use the
optimized path for the case when no type conversion is
necessary. The library doesn't allow this path to be
unregistered although the application can redefine it as often
as it likes.
Fixed the type conversion tables in preparation for MT-safety
and to fix previosly-known design bugs wrt. unregistering
conversion functions or changing the C function associated
with a conversion path. The MT-safety stuff is documented in a
separate white paper.
Increased the conversion function debugging name from 9
characters to 31 characters so the output can be more
descriptive.
Moved conversion path statistics from the H5T_cdata_t member
into the conversion path itself. This makes H5T_cdata_t
contain only application-visible data structures.
./src/H5A.c
./src/H5D.c
./src/H5Ofill.c
./src/H5P.c
./src/H5T.c
Improved the way type conversion functions are called so the
caller doesn't have to check for data type debugging and
increment type conversion timers and statistics.
Changed check for no-op conversion since it is now
application-definable and there may even be more than one
definition at a time in a multi-threaded application (one
thread might be using the no-op conversion path when some
other thread changes its definition -- the first thread still
sees the original defintion until it's done with the
operation).
./doc/html/Datatypes.html
Updated the user guide to reflect the changes to data type
conversion registration functions.
./bin/trace
./src/H5.c
Added tracing support for the new H5T_pers_t data type.
./test/dtypes.c
Added printf to display alignment value if non-aligned data
types are being tested.
./test/h5test.c
Modified the H5Tunregister() calls to use the new
arguments. All 94 of those calls can be replaced by a single
call to the new H5Tunregister() function.
./src/H5.c
Added HAVE_GETTIMEOFDAY around an auto variable and
initialized the elapsed time to zero when gettimeofday() is
not available.
./src/H5F.c
./src/H5Fprivate.h
./src/H5P.c
Added an H5*_init() functions which do the same thing as
H5*_init_interface() but can be called from other packages and
don't do anything if the interface is already initialized.
This fixes a couple memory leaks in applications that
repeatedly close and open the library with H5close().
./src/H5Tconv.c
Optimized some data alignment code in the hardware conversion
functions.
Hardware conversions accumulate statistics about source and
destination data alignment and print that information when the
conversion function is unregistered (usually when the
application exits) if data type debugging is compiled into the
library and enabled at run-time.
The conversion caching was cleaned up for the compound data
type conversion function. It now caches conversion paths in a
manner that will be MT-safe and is much simpler than the old
method. Also cleaned up some array index maps.
./src/H5detect.c
Fixd mispelling of alingemnt.
./src/H5private.h
Changed `TRUE' to `1' in assignment to interface_initialize_g
in FUNC_ENTER macro definition.
./tools/testh5dump.sh
Completely rewritten to make it shorter, better documented,
and conforming to most of the other test outputs.
The comparison of the actual output with the expected output
is insensitive to differences in white space. The test now
passes for the first time on Linux where the output width
wasn't as expected but the output was otherwise correct.
./tools/testfiles/tall-1.ddl
./tools/testfiles/tall-2.ddl
./tools/testfiles/tall-3.ddl
./tools/testfiles/tattr-1.ddl
./tools/testfiles/tattr-2.ddl
./tools/testfiles/tattr-3.ddl
./tools/testfiles/tattr-4.ddl
./tools/testfiles/tcomp-1.ddl
./tools/testfiles/tcomp-2.ddl
./tools/testfiles/tcomp-3.ddl
./tools/testfiles/tcomp-4.ddl
./tools/testfiles/tdset-1.ddl
./tools/testfiles/tdset-2.ddl
./tools/testfiles/tdset-3.ddl
./tools/testfiles/tdset-4.ddl
./tools/testfiles/tgroup-1.ddl
./tools/testfiles/tgroup-2.ddl
./tools/testfiles/tgroup-3.ddl
./tools/testfiles/tlink-1.ddl
./tools/testfiles/tlink-2.ddl
./tools/testfiles/tlink-3.ddl
./tools/testfiles/tlink-4.ddl
./tools/testfiles/tlink-5.ddl
./tools/testfiles/tslink-1.ddl
./tools/testfiles/tslink-2.ddl
Changed `../h5dump' to just `h5dump'.
./config/alpha-dec-osf4.0
Added more warning and optimization switches to the native
compiler.
H5F_istore_lock and H5F_istore_inlock to prevent race between reading and
writing data chunks that caused "holes" (i.e., sequences of 0s)
in chunked datasets.
----------------------
./INSTALL.parallel [NEW]
We're beginning to unify some of the parallel installation
steps. This file will contain general information for
installing the parallel library. It's not complete yet.
./configure.in
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
Check for xdr_int() in libnsl required on Solaris when linking
with hdf4. It's found on the Irix system I tested which
complains that `-lnsl' didn't resolve any symbols. Oh well.
Fixed the order of searching for libdf and libmfhdf for hdf4
linking.
./configure.in
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
./src/H5Z.c
Check for compress() in libz in order to find older versions
of the library that will still work for hdf4. Added a
separate check for compress2() that hdf5 will use.
./configure.in
./configure [REGENERATED]
./src/H5config.h.in [REGENERATED]
./src/H5.c
./src/H5private.h
./src/H5A.c
./src/H5B.c
./src/H5Bprivate.h
./src/H5D.c
./src/H5F.c
./src/H5Farray.c
./src/H5Fcore.c
./src/H5Ffamily.c
./src/H5Fistore.c
./src/H5Flow.c
./src/H5Fmpio.c
./src/H5Fprivate.h
./src/H5Fsec2.c
./src/H5Fsplit.c
./src/H5Fstdio.c
./src/H5Gent.c
./src/H5Gnode.c
./src/H5Gpkg.h
./src/H5Gprivate.h
./src/H5HG.c
./src/H5HL.c
./src/H5O.c
./src/H5Oattr.c
./src/H5Ocomp.c
./src/H5Ocont.c
./src/H5Odtype.c
./src/H5Oefl.c
./src/H5Ofill.c
./src/H5Olayout.c
./src/H5Omtime.c
./src/H5Oname.c
./src/H5Oprivate.h
./src/H5Osdspace.c
./src/H5Oshared.c
./src/H5Ostab.c
./src/H5R.c
./src/H5RA.c
./src/H5Sall.c
./src/H5Shyper.c
./src/H5Snone.c
./src/H5Spoint.c
./src/H5Sprivate.h
./src/H5Sselect.c
./src/H5T.c
./src/H5Tbit.c
./src/H5Tconv.c
./src/H5Tpkg.h
./src/H5V.c
./test/bittests.c
./test/gheap.c
./test/hyperslab.c
./test/istore.c
./test/tmeta.c
./test/trefer.c
./test/tselect.c
./tools/h5debug.c
./tools/h5tols.c
Added checks for Posix.1g types like `int8_t'. If not defined
then H5private.h defines them. Changed all `int8' etc. to
`int8_t'.
./src/H5A.c
./src/H5D.c
./src/H5F.c
./src/H5G.c
./src/H5I.c
./src/H5P.c
./src/H5R.c
./src/H5RA.c
./src/H5S.c
./src/H5T.c
./src/H5TB.c
./src/H5Z.c
Calling H5*_term_interface() resets interface_initialize_g to
FALSE so a subsequent call to H5open() (implied or explicit)
reinitializes global variables properly.
./src/H5private.h
./src/H5Oefl.c
./src/H5S.c
Changed MAX_SIZET, MAX_SSIZET, MAX_HSIZET, and MAX_HSSIZET to
SIZET_MAX, SSIZET_MAX, HSIZET_MAX, and HSSIZE_MAX to they
match the Posix.1 constants in <limits.h>.
./src/H5T.c
./src/H5Tconv.c
./src/H5Tpkg.h
./src/H5Tprivate.h
./src/H5detect.c
Added 36 more integer hardware conversion functions to the
type conversion table for conversions to/from `long long' and
`unsigned long long'. The `long long' names will be changed
shortly to make them portable to Win32.
Changed H5T_init() to H5T_native_open() and added an
H5T_native_close() to open and close the predefined native
data types.
Increased the initial size of the type conversion table from
64 to 128 entries.
Reordered the 90 new integer conversion functions so the names
that are printed favor `int' over `short' or `long' when two
of them are the same.
./test/dtypes.c
Added hardware and software integer conversion tests for the
56 functions I added recently but not the additional 36
checked in this time. That will come next.
Call H5close() after each test so type conversion statistics
are easier to follow. Try this: $ HDF5_DEBUG=t ./dtypes
Added more debugging output for when things go wrong.
./src/H5private.h
Removed trailing carriage-returns inserted by broken operating
system ;-)
----------------------
./bin/trace
./src/H5.c
./src/H5F.c
./src/H5Fprivate.h
./src/H5Fpublic.h
Added a `scope' argument to H5Fflush() which should be either
H5F_SCOPE_LOCAL or H5F_SCOPE_GLOBAL and determines which files
are flushed (just the specified file or the entire virtual
file).
./src/H5F.c
Added reference counts to the H5F_t struct so we get the
correct behavior between H5Funmount() and H5Fclose().
./src/H5O.c
Fixed a memory leak that happens during error handling.
./test/cmpd_dset.c
./test/unlink.c
Fixed a memory leak.
./test/mount.c
Enabled the H5Fclose() test.
----------------------
./src/H5.c
Fixed a signed vs. unsigned comparison.
./src/H5D.c
Setting a fill value of all zeros will cause the fill value to
be written to the dataset instead of relying on the low-level
file driver initializing unwritten areas with zero.
./src/H5D.c
./src/H5F.c
./src/H5Fprivate.h
./src/H5G.c
./src/H5Gpkg.h
./src/H5Gprivate.h
./src/H5O.c
./src/H5T.c
More file mounting stuff.
./src/H5I.c
Fixed a bug where trying to close an invalid object id caused
a core dump. For instance, H5Gclose(-1).
./MANIFEST
./test/Makefile.in
./test/mount.c [NEW]
Mounting tests.
./src/H5R.c
Fixed a couple (herr_t)NULL casts.
----------------------
./src/H5F.c
./src/H5Fprivate.h
./src/H5G.c
./src/H5Gpkg.h
./src/H5Gprivate.h
./src/H5Gpublic.h
./test/istore.c
./test/tstab.c
Removed CWG functionality.
./tools/Makefile.in
Added `_test' where I added `test' last week. Maybe now the
snapshots will start up again...
----------------------
./doc/html/H5.format.html
./src/H5HG.c
Fixed a bug in the global heap that caused H5HG_read() to
write past the end of the buffer in certain cases.
./test/big.c
The test is skipped if hdf5 was configured with
`--disable-hsizet'.
./src/H5Ofill.c
Data type conversions are implemented for the fill value.
./src/H5.c
Tracing prints one of H5P_FILE_CREATE, H5P_FILE_ACCESS,
H5P_DATASET_CREATE, H5P_DATASET_XFER, or H5P_MOUNT instead of
the more cryptic H5I_TEMPLATE_* constants.
./src/H5D.c
Removed prototype for H5D_find_name().
./src/H5I.c
The GROUP_MASK and ID_MASK are both calculated from GROUP_BITS
instead of being set by hand.
We don't use the sign bit of hid_t; all valid hid_t values are
positive so we can say things like `if ((file=H5Fopen(...))<0)'.
Changed `(int)pow(2.0,x)' to `1<<x' so we don't have to worry
about rounding.
Fixed H5I_get_type() so it doesn't always fail an assertion.
./src/H5E.c
./src/H5Epublic.h
Added minor error H5E_MOUNT
./src/H5F.c
./src/H5Fprivate.h
Added H5Fmount() and H5Funmount(). Mounting and unmounting
works as documented but some of the other things aren't
implemented yet, the biggest being current working groups
always acting on the root of the mount tree, and H5Fclose()
closing the entire tree. The rest of the stuff will be added
shortly...
./src/H5P.c
./src/H5Ppublic.h
Added the H5P_MOUNT property list but haven't implemented any
particular properties for it yet.
./src/H5Gstab.c
Hard links across files return an error instead of failing an
assertion.
----------------------
./src/H5D.c
./src/H5Fistore.c
./src/H5Fprivate.h
Fixed a couple things for parallel hdf5 that I broke with the
last checkin.
./src/H5Ofill.c
If no fill value is specified it isn't added to the object
header.
./tools/h5ls.c
Added a line-feed after `Unknown object' is printed.
./tools/h5tools.c
Fixed printing of scalar values. Need to still track down a
failed assertion when H5Sselect_hyperslab() is called with a
zero size....
